tjgalvin
commented
1 month ago Add a hot fix here with 309cc51
The larger question is why this behavior has started to show itself again. I want to say some strange platform issue with petrichor but will look into later.
The alternative approach to overwriting like I am here is to just return the parset path.
I've noticed in some cases in the prefect continuum flow that the linmos task can be rerun by prefect should it detect something went wrong.
In such a case there is an aserting statement to guarantee that the parset files does not exist. It is becoming a hindrance more than help
It shall be deleted